Guidelines for reporting action simulation studies (GRASS): Proposals to improve reporting of research in motor imagery and action observation.

Researchers from multiple disciplines have studied the simulation of actions through motor imagery, action observation, or their combination. Procedures used in these studies vary considerably between research groups, and no standardized approach to reporting experimental protocols has been proposed. This has led to under-reporting of critical details, impairing the assessment, replication, synthesis, and potential clinical translation of effects. We provide an overview of issues related to the reporting of information in action simulation studies, and discuss the benefits of standardized reporting. We propose a series of checklists that identify key details of research protocols to include when reporting action simulation studies. Each checklist comprises A) essential methodological details, B) essential details that are relevant to a specific mode of action simulation, and C) further points that may be useful on a case-by-case basis. We anticipate that the use of these guidelines will improve the understanding, reproduction, and synthesis of studies using action simulation, and enhance the translation of research using motor imagery and action observation to applied and clinical settings.

What we imagine learning from watching others: how motor imagery modulates competency perceptions resulting from the repeated observation of a juggling action.

Although motor learning can occur from observing others perform a motor skill (action observation; AO), observers' confidence in their own ability to perform the skill can be falsely increased compared to their actual ability. This illusion of motor competence (i.e., 'over-confidence') may arise because the learner does not gain access to sensory feedback about their own performance-a source of information that can help individuals understand their veridical motor capabilities. Unlike AO, motor imagery (MI; the mental rehearsal of a motor skill) is thought to be linked to an understanding of movement consequences and kinaesthetic information. MI may thus provide the learner with movement-related diagnostic information, leading to greater accuracy in assessing ability. The present study was designed to evaluate the effects of MI when paired with AO in assessments of one's own motor capabilities in an online observation task. Two groups rated their confidence in performing a juggling task following repeated observations of the action without MI (OBS group; n = 45) or with MI following observation (OBS+MI; n = 39). As predicted, confidence increased with repeated observation for both groups, yet increased to a greater extent in the OBS relative to the OBS+MI group. The addition of MI appeared to reduce confidence that resulted from repeated AO alone. Data support the hypothesis that AO and MI are separable and that MI allows better access to sensory information than AO. However, further research is required to assess changes in confidence that result from MI alone and motor execution.

Disrupting somatosensory processing impairs motor execution but not motor imagery.

While motor imagery (MI) is thought to be 'functionally equivalent' with motor execution (ME), the equivalence of feedforward and feedback mechanisms between the two modalities is unexplored. Here, we tested the equivalence of these mechanisms between MI and ME via two experiments designed to probe the role of somatosensory processing (Exp 1), and cognitive processing (Exp 2). All participants were engaged in a previously established force-matching task adapted for MI. A reference force was applied (on scale of 1-10, with higher numbers indicative of greater force) to one index finger while participants matched the force with their opposite index finger via ME or MI (control conditions). Participants then rated the force on the same scale of 1-10. Exp 1: Participants (N = 27) performed the task with tactile stimulation (ME+TAC, MI+TAC) in addition to control conditions. Exp 2: Participants (N = 26) performed the task in dual-task conditions (ME+COG, MI+COG) in addition to control conditions. Results indicate that (Exp 1) tactile stimulation impaired performance in ME but not MI. Dual-task conditions (Exp 2) were not shown to impair performance in either practice modality. Findings suggest that while somatosensory processing is critical for ME, it is not for MI. Overall we indicate a functional equivalence between feedforward/back mechanisms in MI and ME may not exist.

Examining the role of the supplementary motor area in motor imagery-based skill acquisition.

Motor imagery (MI) and physical practice (PP) have been seen as parallel processes that can drive acquisition of motor skills. Emerging evidence, however, suggests these two processes may be fundamentally different, whereby MI-based motor skill acquisition relies more on effector-independent encoding of movement relative to PP. This alternate view is supported by evidence where real and virtual lesions to brain areas involved in visuospatial processing impair MI-based skill acquisition, and via behavioural studies showing perceptual, but not motor, transfer impairs skill acquisition via MI whereas this effect is reversed in PP. This study further investigated the degree to which MI utilizes effector-independent encoding of movement by investigating the role of the supplementary motor area (SMA), an area involved in perceptual to motor transformations, in MI-based motor skill acquisition. Sixty-four participants completed a serial reaction time paradigm following assignment to one of four groups based on training modality (MI or PP) and stimulation type (sham stimulation or continuous theta burst stimulation to inhibit the SMA). Faster reaction times (RTs) to elements of a repeated sequence in comparison to randomly generated elements indicated that sequence-specific learning occurred. Learning occurred in both PP and MI, with the magnitude of learning significantly smaller in MI. Inhibitory stimulation impaired learning in both modalities. In the context of a framework that distinguishes effector-independent and -dependent components of learning, these findings indicate the SMA plays a role in developing motor chunks in both PP and MI facilitating effector-independent learning in both modalities.

Probing the temporal dynamics of movement inhibition in motor imagery.

Beyond the lack of overt movement in motor imagery (MI), MI is thought to be functionally equivalent to motor execution (ME). Two theories appear viable to explain the neural mechanism underlying the inhibition of movement in MI, with one suggesting the inhibition of movement in MI occurs early in the planning process, and the other suggesting it occurs after the planning for movement is compete. Here we sought to generate evidence related to the timing of movement inhibition in MI. Participants performed a motor task via MI and ME that had distinct preparation and performance phases, with brain activity obtained throughout. Analysis of sensor-level data was performed to isolate event related desynchrony (ERD) in the mu and beta frequency bands in both a sensorimotor and left parietal region of interest (ROI). The magnitude of ERD in the sensorimotor ROI was significantly greater in ME than MI during both the preparatory and performance phases. The reduced ERD in the mu and beta frequency bands in the sensorimotor ROI during the preparatory phase for MI, compared to ME, suggests that movement planning is inhibited (or at least reduced) in MI, contributing to the lack of movement. While past work has shown that the networks of functional brain activity underlying MI and ME are heavily overlapping, differences in the temporal dynamics of this activity suggest that MI and ME are not equivalent processes.

Disruption of motor imagery performance following inhibition of the left inferior parietal lobe.

The left inferior parietal lobe (IPL), a brain region localized to the ventro-dorsal stream, is known to be critical to motor imagery (MI) performance. Yet its specific role in processes underlying MI, namely the generation, maintenance, manipulation, and controllability of motor images, is conflicting in the literature. To determine the specific role of the left IPL in MI, the current study sought to examine the effect inhibition of the left IPL has on performance on two disparate measures thought to probe different MI processes within the same participants. Participants (N = 31) completed the hand laterality judgment task (HLJT), employed to probe processes related to manipulation and controllability, and mental chronometry, employed to probe processes related to generation and maintenance, after receiving either inhibitory transcranial magnetic stimulation to the left IPL (Active-TMS group), or with the coil angled away from the scalp (Sham group). Impaired performance on the HLJT was observed following active TMS relative to sham. Similar mental chronometry performance resulted regardless of left IPL inhibition. In showing that inhibition of the left IPL selectively disrupted performance on the HLJT but not mental chronometry, our findings indicate that the left IPL is specifically involved in image manipulation and controllability during MI. Ultimately, the current study extends our understanding of the role of the left IPL in MI.

Experience modulates motor imagery-based brain activity.

Whether or not brain activation during motor imagery (MI), the mental rehearsal of movement, is modulated by experience (i.e. skilled performance, achieved through long-term practice) remains unclear. Specifically, MI is generally associated with diffuse activation patterns that closely resemble novice physical performance, which may be attributable to a lack of experience with the task being imagined vs. being a distinguishing feature of MI. We sought to examine how experience modulates brain activity driven via MI, implementing a within- and between-group design to manipulate experience across tasks as well as expertise of the participants. Two groups of 'experts' (basketball/volleyball athletes) and 'novices' (recreational controls) underwent magnetoencephalography (MEG) while performing MI of four multi-articular tasks, selected to ensure that the degree of experience that participants had with each task varied. Source-level analysis was applied to MEG data and linear mixed effects modelling was conducted to examine task-related changes in activity. Within- and between-group comparisons were completed post hoc and difference maps were plotted. Brain activation patterns observed during MI of tasks for which participants had a low degree of experience were more widespread and bilateral (i.e. within-groups), with limited differences observed during MI of tasks for which participants had similar experience (i.e. between-groups). Thus, we show that brain activity during MI is modulated by experience; specifically, that novice performance is associated with the additional recruitment of regions across both hemispheres. Future investigations of the neural correlates of MI should consider prior experience when selecting the task to be performed.

Motor imagery-based implicit sequence learning depends on the formation of stimulus-response associations.

Implicit sequence learning (ISL) occurs without conscious awareness and is critical for skill acquisition. The extent to which ISL occurs is a function of exposure (i.e., total training time and/or sequence to noise ratio) to a repeated sequence, and thus the cognitive mechanism underlying ISL is the formation of stimulus-response associations. As the majority of ISL studies employ paradigms whereby individuals unknowingly physically practice a repeated sequence, the cognitive mechanism underlying ISL through motor imagery (MI), the mental rehearsal of movement, remains unknown. This study examined the cognitive mechanisms of MI-based ISL by probing the link between exposure and the resultant ISL. Seventy-two participants underwent MI-based practice of an ISL task following randomization to one of four conditions: 4 training blocks with a high (4-High) or low (4-Low) sequence to noise ratio, or 2 training blocks with a high (2-High) or low (2-Low) sequence to noise ratio. Reaction time differences (dRT) and effect sizes between repeated and random sequences assessed the extent of learning. All groups showed a degree of ISL, yet effect sizes indicated a greater degree of learning in groups with higher exposure (4-Low and 4-High). Findings indicate that the extent to which ISL occurs through MI is impacted by manipulations to total training time and the sequence to noise ratio. Overall, we show that the extent of ISL occurring through MI is a function of exposure, indicating that like physical practice, the cognitive mechanisms of MI-based ISL rely on the formation of stimulus response associations.

The effector independent nature of motor imagery: Evidence from rTMS induced inhibition to the primary motor cortices.

Motor imagery (MI), the mental rehearsal of movement, facilitates learning by driving brain activation similar to that of physical practice (PP). However, a growing body of evidence suggests that learning via MI relies more on effector independent as opposed to effector dependent encoding. One approach to probing the nature of MI based learning is to study the primary motor cortex (MC), a brain region known to be critical to effector dependent encoding, but whose involvement in MI is debatable. The current study sought to inform on the nature of MI-based learning by examining the extent to which participants could learn via MI following inhibition of the MC using repetitive transcranial magnetic stimulation (TMS). Forty-seven participants completed an MI-based implicit sequence learning paradigm after receiving inhibitory TMS to the contralateral or ipsilateral MC (TMS groups), or with the coil angled away from the scalp (Sham). The extent to which participants learned was assessed via reaction time differences (dRT) and effect sizes between repeated and random sequences. Similar dRT values and moderate effect sizes were observed across all groups, providing evidence that inhibition of the MC did not disrupt MI-based learning. As the MC is critical to effector dependent encoding, the current findings suggest that MI-based learning does not rely on effector dependent encoding and unlike PP, is more effector independent in nature. Ultimately, these results inform on the nature of MI-based learning.

Skill acquisition via motor imagery relies on both motor and perceptual learning.

Motor imagery (MI), the mental rehearsal of movement, is an effective means for acquiring a novel skill, even in the absence of physical practice (PP). The nature of this learning, be it perceptual, motor, or both, is not well understood. Understanding the mechanisms underlying MI-based skill acquisition has implications for its use in numerous disciplines, including informing best practices regarding its use. Here we used an implicit sequence learning (ISL) task to probe whether MI-based skill acquisition can be attributed to perceptual or motor learning. Participants (n = 60) randomized to 4 groups were trained through MI or PP, and were then tested in either perceptual (altering the sensory cue) or motor (switching the hand) transfer conditions. Control participants (n = 42) that did not perform a transfer condition were utilized from previous work. Learning was quantified through effect sizes for reaction time (RT) differences between implicit and random sequences. Generally, PP-based training led to lower RTs compared with MI-based training for implicit and random sequences. All groups demonstrated learning (p < .05), the magnitude of which was reduced by transfer conditions relative to controls. For MI-based training perceptual transfer disrupted performance more than for PP. Motor transfer disrupted performance equally for MI- and PP-based training. Our results suggest that MI-based training relies on both perceptual and motor learning, while PP-based training relies more on motor processes. These results reveal details regarding the mechanisms underlying MI, and inform its use as a modality for skill acquisition. (PsycINFO Database Record

Characterizing skill acquisition through motor imagery with no prior physical practice.

Motor learning depends upon plasticity in neural networks involved in the planning and execution of movement. Physical practice (PP) is the primary means of motor learning, but it can be augmented with nonphysical forms of practice including motor imagery (MI)-the mental rehearsal of movement. It is unknown if MI alone, without prior PP of a movement, can produce robust learning. Here the authors used an implicit sequence learning task to explore motor learning via MI alone or PP. Participants underwent implicit sequence learning training via MI (n = 31) or PP (n = 33). Posttraining reaction time was faster for implicit versus random sequences for both the MI group (M = 583 ± 84 ms; 632 ± 86 ms, d = 0.59) and PP group (M = 532 ± 73 ms; 589 ± 70 ms, d = 0.80), demonstrating that MI without PP facilitated skill acquisition. Relative to MI alone, PP led to reduced reaction time for both random (d = 0.65) and implicit sequences (d = 0.55) consistent with a nonspecific motor benefit favoring PP over MI. These results have broad implication for theories of MI and support the use of MI as a form of practice to acquire implicit motor skills. (PsycINFO Database Record

Motor imagery-based skill acquisition disrupted following rTMS of the inferior parietal lobule.

Motor imagery (MI), the mental rehearsal of motor tasks, has promise as a therapy in post-stroke rehabilitation. The potential effectiveness of MI is attributed to the facilitation of plasticity in numerous brain regions akin to those recruited for physical practice. It is suggested, however, that MI relies more heavily on regions commonly affected post-stroke, including left hemisphere parietal regions involved in visuospatial processes. However, the impact of parietal damage on MI-based skill acquisition that underlies rehabilitation remains unclear. Here, we examine the contribution of the left inferior parietal lobule (IPL) to MI using inhibitory transcranial magnetic stimulation (TMS) and an MI-based implicit sequence learning (ISL) paradigm. Participants (N = 27) completed the MI-based ISL paradigm after receiving continuous theta burst stimulation to the left IPL (TMS), or with the coil angled away from the scalp (sham). Reaction time differences (dRT) and effect sizes between implicit and random sequences assessed success of MI-based learning. Mean dRT for the sham group was 36.1 ± 28.2 ms (d = 0.71). Mean dRT in the TMS group was 7.7 ± 38.5 ms (d = 0.11). These results indicate that inhibition of the left IPL impaired MI-based learning. We conclude that the IPL and likely the visuospatial processes it mediates are critical for MI performance and thus MI-based skill acquisition or learning. Ultimately, these findings have implications for the use of MI in post-stroke rehabilitation.